Headline

Assistant Vice President @Hitachi | Production RAG & LLM Apps | Agents · LLM Guardrails | Forward Deployed AI Engineer (FDE)

About

I build production-grade Generative AI for banks and fintechs — systems that survive real users, real volumes, and real auditors. 12+ years across banking and financial services (currently AVP at Hitachi Payment Services; earlier at Barclays). I take RAG and LLM applications from prototype to production on regulated data — loan agreements, KYC files, regulatory circulars — with the guardrails, evaluation, and observability that compliance teams expect. What I deliver: • RAG & Document Q&A — citation-grounded answers over your internal documents. Chunking, metadata filtering, and re-ranking tuned for domain accuracy. (Cut manual review ~45% on a 2,000+ document banking corpus.) • LLM Guardrails & Evaluation — responsible-AI controls, eval harnesses, and observability so you can trust outputs and prove it. • AI/ML & NLP — complaint categorization, transaction intelligence, document processing at scale • POC → Production — I don't stop at a demo. Containerized, monitored, CI/CD, built to scale. Stack: Python, LangChain/LlamaIndex, OpenAI & Hugging Face, vector DBs (FAISS, Pinecone, ChromaDB, Weaviate), PostgreSQL, AWS/GCP/Azure, Docker, Kafka.

Location

Pune/Pimpri-Chinchwad Area

AI stack

antigravity

opencode

claudecode

Work experience

Assistant Vice President at Hitachi Payment Services Pvt. Ltd., Feb 2025 - Present · 1 yr 7 mos, Mumbai, Maharashtra, India, I build production-grade Generative AI for banks and fintechs — systems that survive real users, real volumes, and real auditors. I take RAG and LLM applications from prototype to production on regulated data — loan agreements, KYC files, regulatory circulars — with the guardrails, evaluation, and observability that compliance teams expect. What I deliver: • RAG & Document Q&A — citation-grounded answers over your internal documents. Chunking, metadata filtering, and re-ranking tuned for domain accuracy. (Cut manual review ~45% on a 2,000+ document banking corpus.) • LLM Guardrails & Evaluation — responsible-AI controls, eval harnesses, and observability so you can trust outputs and prove it. • Agentic AI Workflows — multi-step automations that take action, not just chat. • AI/ML & NLP — complaint categorization, transaction intelligence, document processing at scale (100K+ tx/day in production). • POC → Production — I don't stop at a demo. Containerized, monitored, CI/CD, built to scale. Stack: Python, LangChain/LlamaIndex, OpenAI & Hugging Face, vector DBs (FAISS, Pinecone, ChromaDB, Weaviate), PostgreSQL, AWS/GCP/Azure, Docker, Kafka.

Vice President of Artificial Intelligence at Hitachi Payment Services Pvt. Ltd., Feb 2025 - Present · 1 yr 7 mos, Architected a RAG-based document Q&A system over 2K+ banking documents (loan agreements, KYC files, regulatory circulars) on LangChain + OpenAI + FAISS, cutting manual review time by ~45%. Engineered chunking strategies, metadata filtering, and re-ranking to lift retrieval precision on domain-specific financial queries. Built an ML/NLP classification system in Python to auto-categorize customer complaints; designed evaluation harnesses and observability hooks. Architected a microservices payment processing platform (SpringBoot + .NET Core) on PostgreSQL and AWS (EC2, S3, RDS, Lambda, CloudWatch), serving 1M+ transactions/day. Implemented event-driven architecture on Apache Kafka for real-time transaction processing and data streaming. Delivered React + Next.js + TypeScript front-ends and .NET Core REST APIs with secure authN/authZ. Set up CI/CD on GitHub Actions and AWS-native pipelines; cut application load times by 40%. Containerized services with Docker; ran knowledge-sharing sessions on ML, microservices, and DevOps.

Lead Application Developer at i-TEK RFID, Aug 2023 - Nov 2024 · 1 yr 4 mos, Pune District, Maharashtra, India, Team Leadership & Management: Directed a cross-functional team of 4 engineers to deliver a cloud-native platform, resulting in 99.9% uptime Managed full-stack development team of 8 developers, delivering projects 25% ahead of schedule. Led agile transformation initiative for 9-person engineering team, increasing sprint velocity by 45% Process & Performance Improvements: Implemented DevOps practices and infrastructure-as-code, cutting deployment time from 4 hours to 15 minutes Optimized database queries and caching strategies, improving application response time by 60% Introduced code quality standards and peer review process, reducing production incidents by 75%

Senior Application Specialist at Infotek Software Ltd, Jul 2023 - Sep 2024 · 1 yr 3 mos, Led full-stack delivery on enterprise applications spanning Python services, .NET Core APIs, and React front-ends. Drove architecture, design reviews, and code quality across the team; mentored engineers on microservices and async/event-driven patterns. Owned CI/CD on Azure DevOps and GitHub Actions; introduced Docker-based deployments and Prometheus/Grafana observability.

Lead Consultant at Magic EdTech, Jul 2022 - Jun 2023 · 1 yr, Remote

Technical Lead at Magic Software Inc., Jul 2022 - Jun 2023 · 1 yr, Tech-led a multi-team product squad shipping production features on .NET Core and React. Designed REST APIs, set coding standards, and ran code reviews; introduced unit-test discipline and CI checks.

Education

Master of Technology (M.Tech.), Computer Science and Engineering, Walchand College of engineering, Sangli

SSC, Shanta Prabhu Parasla, Ajansond